Envisioning Artificial Life

For billions of years Earth has been filled with a great many small biological organisms. Each one is designed to not die, and to collect and process enough resources to make close but not exact copies of itself, copies which then must each search for and defend a niche wherein this story can be repeated.

Recently humans have introduce some new kinds of entities, including human organizations and artificial devices. It seems obvious to me that within a millennia all these things will merge into something new: “artificial life”. Which will then forever more dominate the universe (if it survives). And it seems worth trying to think through what this will look like. Here are eight key features I predict for artificial life:

Advanced – Over billions of years bio life has developed some amazing innovations, and used these to colonize a wide range of niches on Earth. But there are still many places near here where life has had at most only a mild impact. And in far less than billions of years humans have found a great many innovations that bio never found, or made much use of. Such as metal smelting, trains, jet engines, nuclear power, rocket ships, radio, computers, and far far more. Using these, humans have been able to go more places, use more kinds of energy, and do more things. Artificial life will mix up these bio- and human-discovered design elements in a huge variety of ways, and colonize a far wider range of niches. Artificial life will live on and in a wide range of planets, stars, rocks, clouds, and volumes, and reorganize such things to make whole new places and things. Artificial life will grow faster, at least until the solar system is nearly filled.

Specialized – Lone bio cells have to do basically everything themselves. Which means they can’t achieve the very large scale and scope economies possible via a division of labor. Multi-cellular organisms have more of an internal specialization, but still each organism must do most everything itself. Social animals allow still more division of labor, but even so only over relatively small scales. In contrast, artificial life can manage a civilization-wide division of labor, limited mainly by transport costs, allowing each particular part to take on very specialized roles. Most artificial creatures can’t forage, digest, reproduce, etc. by themselves very well, but are designed to function well mainly within large complex societies. Artificial life is quite inter-dependent.

Informed – Bio organisms mainly know about what they can directly see, and the insight implicitly embodied in their genes. In contrast, artificial creatures can also use large specialized communication networks and institutions to learn about a great many things. The main limits are data, costs of distance and computation, and strategic incentives to deceive.

Invited – Bio tends to just make stuff when it can, without attending much to if there is a demand for that stuff. In contrast, artificial stuff tends to be made when and where demand is envisioned. New firms are made when investors guess they can make a profit. New couches are made and shipped to stores where firms guess buyers are likely to want them. And so on. Invited creatures have to worry less about their place in the world, or wonder why they exist; they wouldn’t exist unless there had been at least a reasonable place for them.

Designed – New bio things tend to be quite similar to their creators, adding mostly random variations. In contrast, artificial stuff is often more carefully designed to be suitable for some demand, and substantially different from prior things. A building is designed to fit its lot, a firm is designed to fit its market, and so on. Compared to bio designs, artificial designs draw on a much wider range of prior designs, and a lot more effort goes into figuring out these new designs. Artificial life is more like one big civ-wide species, and less breaks into separate lineages with designs only derived from its species. Some artificial creatures specialize in designing parts of other ones.

Governed – Artificial creatures coordinate to avoid destructive conflicts via empowering governance organizations, such as firms, clubs, law, and government. Such governance tends to be of wider scope and more stable than other structures, and is thus harder to influence. Yet each such structure has some ultimate owners who control it over larger scopes of space and time. When interacting with an artificial creature, one may want to know about the larger governance units with which it is allied.

Varied – While before humans biological organisms had a huge range of sizes and abilities, our more recent era has been dominated by creatures with quite similar sizes and abilities. Namely humans. With artificial life variety will return, and greatly expand. Trying to count military or political power by counting heads just won’t work at all.

Owned – The bio world has little property; stuff can and is oft stolen. Only property that can be directly created and defended exists. But the artificial world has far more property (and liability) rights. As a result, to use resources one must create, purchase, or inherit them. Thus someone will have to pay for the resources and design effort needed to make each new piece of artificial life. Those who pay to create something will often reserve some property (or liability) rights over it, often in the form of debt or equity. Though the created thing will usually hold some rights in itself. Creatures and groups will often trade shares in each other as a way to align their interests more closely.

This last owned features seems the most likely to bother people today. But my job is to tell you what seems most likely, not what you want to hear.